Sage Journals: Discover world-class research

Abstract

BACKGROUND:

Although antidepressant medication (ADM) has produced small advantages over pill placebo in randomized controlled trials, consuming ADM has predicted prospectively increasing depressive symptom severity in samples of community-dwelling adults.

OBJECTIVE:

We extended the community literature by testing ADM’s relations to changes in personality and quality of life that may underpin depression.

METHOD:

In this longitudinal, observational study, community-dwelling adults (N = 601) were assessed twice, 8 months apart on average. Assessments included depressive symptoms, personality, life satisfaction and quality, and prescription medication consumption.

RESULTS:

Consuming ADM at time 1 predicted relative increases in depressive symptoms (dysphoria), maladaptive traits (negative affect, negative temperament, disinhibition, low conscientiousness), personality dysfunction (non-coping, self-pathology), and decreases in life satisfaction and quality from time 1 to 2, before and after adjustment for age, gender, race, income, education, physical health problems, and use of other psychotropics. In no analysis did ADM use predict better outcomes.

CONCLUSION:

Among community-dwelling adults, ADM use is a risk factor for psychosocial deterioration in domains including depressive symptoms, personality pathology, and quality of life. Until mechanisms connecting ADM to poor outcomes in community samples are understood, additional caution in use of ADM and consideration of empirically supported non-pharmacologic treatments is prudent.

Keywords

Antidepressant medications personality quality of life risk community

1. Introduction

What happens over time to adults who consume antidepressant medication (ADM)? Many clinicians, researchers, and patients assert that ADM consumers, compared to similar adults who do not take these medications, experience decreased depressive symptom severity, improved functioning, and greater quality of life. In randomized controlled trials (RCTs), ADM has often produced statistically significant but small, perhaps clinically insignificant, benefits over pill placebo among adults with depression and related anxiety disorders [1–5]. However, as routinely prescribed and consumed in the community, ADM has predicted worsening symptoms in observational studies [6–11].

Effects of ADM could differ between RCTs and community samples for several interrelated reasons. Importantly, ADM effects are typically measured after weeks in RCTs [2,5], whereas community-dwelling adults often consume ADM for years [12,13]. Biochemical processes producing short-term decreases in depressive symptoms during acute treatment with ADM may eventually trigger compensatory reactions. These compensatory reactions, also called opponent processes or oppositional tolerance, may then cause, or contribute to, increases in depressive symptom severity, chronic depression, treatment resistance, withdrawal syndrome, relapse, and recurrence [6,14,15].

When attempting to discontinue ADM, roughly half of patients experience an aversive withdrawal syndrome lasting from days to months or longer [16,17]. Avoidance of this withdrawal syndrome, and misinterpretation of withdrawal as depressive relapse, may lead some consumers and clinicians to continue or restart ADM in the absence of benefits and the presence of other harms [6,18]. Harmful side effects of ADM include gastrointestinal distress, emotional blunting, sleep disturbance, weight gain, and impaired sexual functioning [19,20]. The cumulative burden of withdrawal syndrome and/or continuing side effects may increase depressive symptoms and related problems.

Previous observational studies of community samples provided evidence of potential ADM effects over longer periods. For example, in a U.S. national sample of middle-aged adults with MDD, participants using ADM at baseline had increases in depressive symptoms over 9-year retest intervals compared to adults with MDD who did not use ADM or received other forms of treatment [10]. In a sample of young Swiss adults selected for high psychological distress and followed for up to 30 years, ADM use similarly predicted prospectively increasing depression [7]. Moreover, in a U.S. national sample, consumption of ADM predicted increasing psychological distress (depression and anxiety symptoms) and functional impairment (reduced physical, cognitive, social, and work performance) over a 1-year interval. At the same time, higher distress and dysfunction predicted prospectively increasing use of ADM, suggesting a harmful feedback loop [11]. Finally, psychiatric inpatients consuming ADM and later discharged into the community with a prescription for continued ADM were re-hospitalized more often and longer over the next year than a matched comparison sample not using ADM [8].

Past observational studies of ADM outcomes have focused primarily on depressive symptoms and other psychological distress measures. However, depression extends beyond its defining symptoms [21] and intertwines with personality and quality of life. Personality dimensions especially relevant to depression include both traits and personality dysfunction. Particularly relevant traits include high neuroticism/negative affectivity, low extraversion/high detachment, and high disinhibition/low conscientiousness [22,23]. Relevant personality dysfunction includes problems with self-functioning (e.g., tolerance of distress and frustration, integrated sense of self supporting adaptive behavior) and interpersonal functioning (e.g., successfully fulfilling social roles, maintaining positive social relationships) [24,25]. Moreover, persons with or at risk for depression more often experience dissatisfaction and perceive limitations in their general life circumstances and in specific domains such as health, social relationships, physical functioning, leisure, and work. That is, reduced life satisfaction and quality also relate to depression, both concurrently and longitudinally [26–28].

The current longitudinal observational study clarified the breadth of potential ADM effects among community-dwelling adults, extending previous research by evaluating not only depressive symptoms but also changes in personality pathology and quality of life that may underlie depression. First, we tested the hypothesis that adults consuming (vs. not consuming) ADM display relative increases in depressive symptoms severity, replicating past research. Extending the literature, we also hypothesized that ADM use predicts increases in personality pathology, including both trait and functioning dimensions. Finally, we hypothesized that ADM consumption predicts deterioration in life satisfaction and quality over time.

2. Method

2.1. Participants and recruitment

Data were drawn from a research program designed to improve theory and measurement of personality pathology (see [29,30] for additional study design details, as well as analyses addressing other goals in a subset of the current sample). The study protocol was reviewed and approved by the Institutional Review Board of the University of Notre Dame (approval number 17-12-4289). Data were collected in years 2012–2016. Participants were at least 18 years old, living in the community (e.g., not inpatients or incarcerated), able to respond to interviews and questionnaires in English, and did not have dementia, intellectual disability, or active psychosis. Table 1 summarizes the sample of community-dwelling adults’ background characteristics.

Table 1
Participant characteristics and medication use at time 1

Variable M or % SD ADM r

Antidepressant medication consumption (ADM) 42.8%

Gender

Female 56.7% 0.10

Male 43.3% —

Age in years 45.57 13.34 0.06

Race

Black 21.0% −0.07

Other 10.9% −0.02

White 68.1% —^a

Income level 2.52 1.42 −0.17

Education level 1.89 0.73 0.00

Physical health problems 1.15 0.98 0.22

Participant recruitment source

Mental health outpatients 50.1% 0.42

High risk for personality pathology 49.9% —

Other psychotropic consumption

Antianxiety medication 25.0% 0.37

Antipsychotic medication 15.5% 0.31

Mood stabilizer medication 13.6% 0.18

Sleep medication 23.5% 0.35

Pain medication 22.1% 0.22

Variable	M or %	SD	ADM r
Antidepressant medication consumption (ADM)	42.8%
Gender
Female	56.7%		0.10
Male	43.3%		—
Age in years	45.57	13.34	0.06
Race
Black	21.0%		−0.07
Other	10.9%		−0.02
White	68.1%		—^a
Income level	2.52	1.42	−0.17
Education level	1.89	0.73	0.00
Physical health problems	1.15	0.98	0.22
Participant recruitment source
Mental health outpatients	50.1%		0.42
High risk for personality pathology	49.9%		—
Other psychotropic consumption
Antianxiety medication	25.0%		0.37
Antipsychotic medication	15.5%		0.31
Mood stabilizer medication	13.6%		0.18
Sleep medication	23.5%		0.35
Pain medication	22.1%		0.22

Note. N = 601, except race and income N = 599 and education N = 600. ADM r is the correlation with antidepressant medication consumption. Correlations in bold, p < 0.05, two-tailed. To include all 601 participants, correlations were computed using full information maximum likelihood estimation. — = Reference category when computing correlation for a categorical variable. ^aAn omnibus test of the relation between race and ADM use yielded X ²(2) = 3.33, p = 0.19.

Participants were recruited from two sources: (1) Mental health outpatients were referred by a community mental health center and local practitioners. A few outpatients were also recruited using newsletters, listservs, and mass emails sent to University of Notre Dame faculty, staff, and graduate students, and by word of mouth. (2) Participants at high-risk for personality pathology were recruited via random dialing of both landlines and cell telephone numbers in the greater South Bend, Indiana area. High-risk participants were not receiving treatment from a mental health provider when recruited but gave 2 or more positive responses on the Iowa Personality Disorder Screen (IPDS) [31], a cut-point that has been shown to have good sensitivity and specificity in identifying personality disorder. A few high-risk participants with above-threshold IPDS responses from the study pilot phase [32] were also recruited and re-screened. Participants recruited from the two sources scored similarly on the depression, personality, and quality of life measures analyzed here (e.g., small standardized mean differences ranging from −0.43 to 0.36 at time 1), and we pooled their data for analysis.

2.2. Procedure

The study design was observational and longitudinal. Researchers presented study information to potential participants verbally and in writing, including study procedures, goals, risks, and benefits. After researchers addressed any questions, participants gave written informed consent. Participants completed assessments at a research facility over 1–2 days, depending on their availability (e.g., work and family schedules) and preferences (e.g., two shorter days vs. one longer day of assessment). The assessment battery was the same for all participants. Nearly all participants completed questionnaires alone using a computer; a research team member assisted a few participants (e.g., due to poor eyesight or unfamiliarity with computers). Trained interviewers (research study staff, advanced graduate students, and Ph.D.-level researchers) conducted semi-structured interviews. Participants agreed to return for additional assessment approximately 6 months later. Researchers used multiple methods to attempt to contact all participants who completed the initial assessment.

2.3. Measures

2.3.1. Medication consumption and background variables

Assessed using a questionnaire given at time 1, we coded age, gender, race, income, education, current use of psychotropic medications, and physical health problems. Yearly income was coded: (1) < $10,000; (2) $10,000–19,999; (3) $20,000–39,999; (4) $40,000–59,999; and (5) ≥ $60,000. Formal education was coded: (1) up to high school diploma or equivalent, (2) some college but less than a bachelor’s degree, (3) bachelor’s degree or more.

Participants self-reported current use of prescription psychotropics in five broad classes. Following the question “Do you regularly take any medication prescribed by a doctor?,” six classes of medication were named and examples provided: “Antidepressants (ex: Prozac, Celexa, Wellbutrin, Lexapro, Paxil, Zoloft, Remeron),” “Antianxiety (ex: Buspar, Benzodiazepines),” “Antipsychotics (ex: Zyprexa, Risperdal, Seroquel, Clozaril, Geodon, Haldol),” “Mood Stabilizers (ex: Depakote, Lithium, Tegretol),” “Sleep Aids (ex: Trazadone, Ambien),” “Pain Medications (ex: Demerol, Vicodin, OxyContin).” We coded (1/yes) or (0/no) to reflect current consumption of each type of medication, including ADM.

The physical health problems scale ranged from 0–3 with one point for each affirmative answer: “Do you regularly take any medication prescribed by a doctor” for a “physical condition?”; “Have you had an operation or major physical illness in the past year?”; and “Do you have a physical disability?” The scale’s alpha internal consistency reliability was 0.52 in the current sample, with an average inter-item correlation (0.27) in the psychometrically desirable range [33].

2.3.2. Depression

We measured depressive symptom severity with the MDD symptom count from the Mini-International Neuropsychiatric Interview (MINI) [34] and with the dysphoria and well-being scales from the Inventory of Depression and Anxiety Symptoms–II (IDAS-II) [35]. Trained interviewers administered the MINI, a semi-structured, diagnostic interview adapted with permission of the author to assess DSM-5 symptoms. We analyzed two scales from the self-report IDAS-II: The 10-item dysphoria scale, which assesses negative emotional and cognitive experiences that are core depression symptoms (e.g., depressed mood, anhedonia, worry, worthlessness), and the 8-item well-being scale, which measures positive emotional and cognitive experiences (e.g., pride, optimism, enthusiasm). In the current sample, MDD assessment had high inter-rater reliability (ICC = 0.99), and the dysphoria (0.89) and well-being (0.88) scales had high alpha internal consistency reliability.

2.3.3. Personality pathology

We measured personality pathology using two measures each of three broad trait dimensions (negative affect/temperament, detachment, and disinhibition/low conscientiousness) and two functioning dimensions (self-pathology and interpersonal pathology) relevant to depression.

Trait measures included the Personality Inventory for DSM-5 (PID-5) [36] and the Schedule for Nonadaptive and Adaptive Personality, 2nd edition (SNAP-2) [37]. The PID-5 was developed in conjunction with the Alternative DSM-5 Model of Personality Disorders and assesses trait facets organized in five domains (negative affectivity, detachment, antagonism, disinhibition, psychoticism) using a 4-point scale (very false or often false to very true or often true). The SNAP-2 assesses personality traits across the normal-abnormal range with true-false format items. The SNAP-2’s 15 core scales form three higher order factors, negative affectivity, positive affectivity, and disinhibition, and an additional conscientiousness scale parallels the PID-5 disinhibition scale conceptually, following a “Big Five” trait model [38]. The trait scales analyzed here, PID-5 negative affectivity (0.78), detachment (0.73), and disinhibition (0.75), as well as SNAP-2 negative temperament (0.92), detachment (0.86), and conscientiousness (0.80), demonstrated moderate-high internal consistency reliability in the current sample.

Personality functioning measures included the Measure of Disordered Personality Functioning (MDPF) [39] and the General Assessment of Personality Disorder (GAPD) [40]. We administered a shortened version of the MDPF [41] assessing the core of the non-coping (e.g., fail more often than succeed, cope poorly) and non-cooperativeness (e.g., can be difficult in dealing with others vs. nice, good-hearted, caring) dimensions, with 22 items rated on a general time frame and a 4-point response scale (definitely false to definitely true). On the GAPD, participants rated items on a 5-point response scale (very unlike me to very like me). We analyzed a 15-item self-pathology (e.g., wonder who real me is, powerless to influence what happens to me) and an 11-item interpersonal pathology (e.g., no close relationships, don’t work to cooperate with people) scales [32]. In the current sample, the non-coping (0.89), non-cooperativeness (0.88), self-pathology (0.92), and interpersonal pathology (0.78) scales had moderate-high internal consistency reliability.

2.3.4. Quality of life

We measured quality of life using the Satisfaction with Life Scale (SWLS) [42] and the general satisfaction scale from the WHO Quality of Life-Brief Form (WHOQOL-BREF) [43]. On the SWLS, participants rated five items on 7-point Likert response format and a general time frame (e.g., “The conditions of my life are excellent”), and we analyzed the total score. On the WHOQOL-BREF, participants responded using a past-2-weeks time frame and five-point rating scale; we analyzed a 9-item general satisfaction scale (e.g., “To what extent do you feel your life to be meaningful?”) [32]. In the current sample, the SWLS (0.77) and WHOQOL-BREF general satisfaction scale (0.87) had acceptable internal consistency reliability.

2.4. Sample size and statistical analyses

The study’s target sample size of 600 (300 outpatients plus 300 high-risk participants) was determined by a combination of feasibility and sufficient sample size for stability of correlational and structural analyses, consistent with past recommendations [44,45]. A sample size of 600 generally provides adequate statistical power (≥ 0.80) to detect small associations in correlation and regression analyses (e.g., r or standardized beta ≥ 0.12), using a two-tailed alpha of 0.05 [46].

Some data were missing, primarily due to attrition from the study (see Ns in Tables 1 and 2). The probability of having missing data at time 2 did not differ significantly between participants consuming (18% missing) versus not consuming (17% missing) ADM at time 1, X ²(1) = 0.11, p = 0.74. To include all participants’ data and increase the generalizability of our findings, we conducted correlation and regression analyses using full information maximum likelihood estimation [47]. Ordinary least squares analyses that excluded participants with missing data produced similar results. We modeled changes in each outcome variable (e.g., negative temperament, quality of life) by regressing time 2 observations on time 1 observations of the same variable, ADM consumption, and covariates (e.g., demographics, other medications). Dichotomous variables (e.g., ADM consumption) were coded 1 (present/true) or 0 (absent/false). In this manner, we compared outcomes of adults using ADM to similar adults who did not report using these medications.

Table 2
Depression, personality, and quality of life measures: descriptive statistics and correlations with time 1 consumption of antidepressant medication (ADM)

Time 1 Time 2

Variable M SD ADM r M SD ADM r

Depressive symptom measures

MDD symptom count 3.50 2.73 0.21 3.03 2.73 0.25

IDAS-II dysphoria 2.37 0.89 0.22 2.27 0.93 0.27

IDAS-II well-being 2.69 0.89 −0.07 2.67 0.85 −0.11

Personality trait measures

PID-5 negative affect 1.30 0.62 0.20 1.20 0.60 0.26

PID-5 detachment 1.10 0.57 0.15 1.07 0.59 0.18

PID-5 disinhibition 0.99 0.53 0.16 0.92 0.54 0.26

SNAP-2 negative temperament 17.04 7.28 0.22 15.64 7.68 0.30

SNAP-2 detachment 15.71 6.87 0.02 15.47 6.69 0.07

SNAP-2 conscientiousness 9.90 5.90 −0.16 9.04 5.77 −0.11

Personality functioning measures

MDPF non-coping 1.28 0.67 0.16 1.22 0.66 0.25

MDPF non-cooperativeness 0.70 0.48 −0.03 0.70 0.47 0.01

GAPD self-pathology 1.38 0.83 0.15 1.29 0.84 0.18

GAPD interpersonal pathology 1.13 0.62 0.02 1.11 0.63 0.08

Quality of life measures

Satisfaction with life scale 3.55 1.39 −0.14 3.63 1.39 −0.18

WHO quality of life scale 3.22 0.70 −0.17 3.27 0.73 −0.27

	Time 1	Time 2
Depressive symptom measures
MDD symptom count	3.50	2.73	0.21	3.03	2.73	0.25
IDAS-II dysphoria	2.37	0.89	0.22	2.27	0.93	0.27
IDAS-II well-being	2.69	0.89	−0.07	2.67	0.85	−0.11
Personality trait measures
PID-5 negative affect	1.30	0.62	0.20	1.20	0.60	0.26
PID-5 detachment	1.10	0.57	0.15	1.07	0.59	0.18
PID-5 disinhibition	0.99	0.53	0.16	0.92	0.54	0.26
SNAP-2 negative temperament	17.04	7.28	0.22	15.64	7.68	0.30
SNAP-2 detachment	15.71	6.87	0.02	15.47	6.69	0.07
SNAP-2 conscientiousness	9.90	5.90	−0.16	9.04	5.77	−0.11
Personality functioning measures
MDPF non-coping	1.28	0.67	0.16	1.22	0.66	0.25
MDPF non-cooperativeness	0.70	0.48	−0.03	0.70	0.47	0.01
GAPD self-pathology	1.38	0.83	0.15	1.29	0.84	0.18
GAPD interpersonal pathology	1.13	0.62	0.02	1.11	0.63	0.08
Quality of life measures
Satisfaction with life scale	3.55	1.39	−0.14	3.63	1.39	−0.18
WHO quality of life scale	3.22	0.70	−0.17	3.27	0.73	−0.27

Note. Time 1 N = 601. Time 2 N = 497, except major depressive disorder (MDD) symptom count N = 501. IDAS-II = Inventory of Depression and Anxiety Symptoms expanded version. SNAP-2 = Schedule for Nonadaptive and Adaptive Personality, 2nd edition. PID-5 = Personality Inventory for DSM-5. MDPF = Measure of Disordered Personality Functioning. GAPD = General Assessment Personality Disorder. WHO = World Health Organization. ADM r is the correlation with antidepressant medication consumption. Correlations in bold, p < 0.05, two-tailed. To include all 601 participants, correlations were computed using full information maximum likelihood estimation.

3. Results

About 43% (257/601) of participants reported consuming ADM at time 1. Consuming ADM correlated significantly with gender (female = 47% using ADM, male = 37%), recruitment source (outpatient = 63%, high-risk = 22%), lower income, greater physical health problems, and consuming other psychotropics (see Table 1). Age, race, and education level were not significantly associated with ADM consumption.

As shown in Table 2, most measures of depressive symptoms, personality, and quality of life correlated significantly with time 1 ADM consumption (12/16 measures at time 1; 13/16 measures at time 2). All significant correlations were in the direction of more distress and pathology among ADM consumers. The strength of time 1 ADM’s significant associations with these problems at time 1 (median |r| = 0.16) did not abate over time but instead tended to increase at time 2 (median |r| = 0.25).

We tested whether consuming ADM at time 1 predicted changes in depressive symptoms, personality, and quality of life measures by regressing each time 2 outcome variable on time 1 observations of the same variable plus ADM consumption. Standardized betas for ADM appear in Table 3. Consumption of ADM at time 1 predicted relative increases in depressive symptomatology (MDD symptom count and dysphoria), maladaptive traits (negative affect and temperament, detachment, disinhibition, and low conscientiousness), personality dysfunction (non-coping, self-pathology, and interpersonal pathology), and decreases in life satisfaction and quality. Use of ADM did not predict changes in well-being or non-cooperativeness scales.

Table 3
Prediction of changes in depression, personality, and quality of life from antidepressant medication consumption (ADM)

Standardized beta for time 1 ADM:

Time 2 outcome variable Controlling time 1 scores on the outcome variable Controlling time 1 scores on the outcome variable plus background covariates^a

Depressive symptom measures

Major depressive disorder symptom count 0.13 0.05

IDAS-II dysphoria 0.12 0.10

IDAS-II well-being −0.07 −0.07

Personality trait measures

PID-5 negative affect 0.10 0.08

PID-5 detachment 0.07 0.05

PID-5 disinhibition 0.14 0.12

SNAP-2 negative temperament 0.13 0.08

SNAP-2 detachment 0.05 0.05

SNAP-2 conscientiousness −0.07 −0.10

Personality functioning measures

MDPF non-coping 0.13 0.13

MDPF non-cooperativeness 0.02 0.00

GAPD self-pathology 0.07 0.07

GAPD interpersonal pathology 0.07 0.04

Quality of life measures

Satisfaction with life scale −0.10 −0.10

WHO quality of life scale −0.14 −0.11

	Standardized beta for time 1 ADM:
Depressive symptom measures
Major depressive disorder symptom count	0.13	0.05
IDAS-II dysphoria	0.12	0.10
IDAS-II well-being	−0.07	−0.07
Personality trait measures
PID-5 negative affect	0.10	0.08
PID-5 detachment	0.07	0.05
PID-5 disinhibition	0.14	0.12
SNAP-2 negative temperament	0.13	0.08
SNAP-2 detachment	0.05	0.05
SNAP-2 conscientiousness	−0.07	−0.10
Personality functioning measures
MDPF non-coping	0.13	0.13
MDPF non-cooperativeness	0.02	0.00
GAPD self-pathology	0.07	0.07
GAPD interpersonal pathology	0.07	0.04
Quality of life measures
Satisfaction with life scale	−0.10	−0.10
WHO quality of life scale	−0.14	−0.11

Note. N = 601. IDAS-II = Inventory of Depression and Anxiety Symptoms expanded version. SNAP-2 = Schedule for Nonadaptive and Adaptive Personality, 2nd edition. PID-5 = Personality Inventory for DSM-5. MDPF = Measure of Disordered Personality Functioning. GAPD = General Assessment Personality Disorder. WHO = World Health Organization. Betas were estimated in multiple regression models. Time 1 ADM coded 1 = consuming ADM or 0 = not consuming ADM. Coefficients in bold, p < 0.05, two-tailed. ^aCovariates were age, gender, race, income, education, recruitment source, physical health problems, and consumption of antianxiety, antipsychotic, mood stabilizer, sleep, and pain medications.

We repeated the time-lagged regression models after adding all participant background measures taken at time 1 as covariates, including age, gender, race, income, education, mental health center outpatient status, physical health problems, and consumption of antianxiety, antipsychotic, mood stabilizer, sleep, and pain medications. We included all background measures as covariates to adjust for both significant and smaller differences between ADM consumers and non-consumers at baseline. In these conservative analyses, ADM consumption continued to predict significant increases in dysphoria, negative affect, negative temperament, disinhibition, low conscientiousness, non-coping, and self-pathology, as well as decreases in life satisfaction and quality (see Table 3).

4. Discussion

The current analyses expanded the research base addressing the question, “What happens over time to community-dwelling adults who consume ADM?”, regarding psychosocial outcomes. We found that ADM consumers experienced relative increases in depressive symptom severity, replicating past research, plus increases in personality pathology and decreases in life satisfaction and quality, extending previous findings and supporting our hypotheses. Increases in personality pathology predicted from ADM consumption included the propensity to experience and express negative emotion (negative affect/temperament), irresponsible and careless behavior (disinhibition/low conscientiousness), and impairment in the sense of self that interferes with adaptive behavior (non-coping and self-pathology).

Relative increases in several distinct dimensions of personality pathology suggests that risks to community-dwelling adults consuming ADM are broad. Although personality is usually fairly stable over time, it is not fixed. Mean personality scores change slowly over adulthood (e.g., decreases in neuroticism and disinhibition) in normative samples and more quickly during the first few weeks of mental health treatments in clinical trials (e.g., decreases in neuroticism and increases extraversion) [48,49]. If ADM applied in RCTs provides a “cause-correction” of personality pathology that contributes to depression [50], a parallel hypothesis is that ADM, as used in the community, produces or worsens personality pathology that drives subsequent increases in depressive symptoms. Future research testing this possibility would require more frequent, repeated assessment but could help explain relative increases in depressive symptoms among community-dwelling adults consuming ADM in the current and earlier samples [7,8,10,11].

Quality of life and life satisfaction are high-priority treatment targets [51], often valued more than decreases in depression-specific symptomatology by patients [52], and central in defining full recovery [53]. Consequently, it is particularly troubling that community-dwelling adults consuming ADM experienced lower life satisfaction and quality at time 1 and relative deterioration in these indices an average of 8 months later. In RCTs, ADM has produced small improvements in life satisfaction and quality compared to control conditions [54,55] but arguably too small to be clinically significant [3]. In the current community sample, ADM’s potentially harmful effects on life satisfaction and quality paralleled that observed for depressive symptomatology, perhaps by similar mechanisms (e.g., increases in maladaptive traits and decreases in adaptive behavior).

However, moderators and mechanisms connecting routine consumption of ADM with poorer outcomes are currently unknown. Several possibilities exist. For example, adults who consumed ADM may have had liabilities or diatheses (e.g., biological, psychological, environmental) that accounted for their poorer outcomes, rather than ADM itself harming participants. That is, community-dwelling adults who elected to consume ADM might have pre-treatment characteristics that distinguished them from their non-consuming counterparts. If so, such ADM-linked liabilities must have exceeded those already captured by concurrent scores on outcome variables (e.g., depressive symptom severity, negative affect/temperament), taking other types of psychotropic medications, physical health problems, receiving specialty care as mental health center outpatients, and low socioeconomic status, all of which we controlled in our analyses. Or perhaps liabilities represented by the decision to consume ADM systematically emerged or worsened over the retest interval, a more complex and so less likely mechanism.

Another possible explanation is that ADM itself caused or contributed to relative increases in depressive symptomatology and personality pathology, as well as decreases in life satisfaction and quality. Such a possibility has been a concern for decades, particularly in the context of high depressive relapse/recurrence rates during and after ADM use, loss of efficacy, withdrawal or discontinuation syndrome, and an accumulating burden of ADM side effects [6,14,56]. More generally, the biochemical actions of ADM are incompletely understood but do not correct a “chemical imbalance” of low serotonin once thought to cause depression [57–59]. Studies with comprehensive assessment and longitudinal follow-up lasting years may advance understanding.

A related possibility is that ADM is only helpful under restrictive conditions and/or to a small subset of patients. These conditions and patients may be present more often in RCTs but less often in routine prescription and community use of ADM. For example, RCTs often enroll patients with less comorbidity (e.g., without substance abuse or active suicidality) and provide them with free, short-term care from experts following a written protocol or algorithm. These RCT design features are unlike routine care involving patients with a wider range of illness severity and comorbidity; prescribers with less expertise in psychiatry who may not follow evidence-based protocols; less frequent and comprehensive assessment to inform treatment decisions; and much longer duration use of ADM [9,60].

We have no reason to believe that the current participants consumed antidepressant medication for periods shorter or longer than usual for community-dwelling adults [12,13], but duration of medication use was not assessed in the current study. Consequently, the reliability and validity of our findings would be strengthened if replicated in future studies assessing additional information about antidepressant medication consumption, such as duration, type, dose, and adherence to prescription instructions. For example, if antidepressant medication is a contributor to increases in depressive symptomatology and personality pathology among community-dwelling adults, then one hypothesis is that these outcomes are stronger for persons consuming antidepressants more consistently, at higher doses, and/or for longer periods.

Regardless of the specific nature of the connection between ADM consumption and measured outcomes, the current results replicate and extend findings of low ADM effectiveness in community samples. In none of the current analyses did ADM consumers fare better. If ADM were robustly helpful, analyses should have detected psychosocial improvement more so, and more often, than deterioration. Thus, any unobserved ADM benefits must have been some combination of rare, weak, and brief. Until the moderators and/or mechanisms connecting community-dwelling adults’ ADM consumption with increasing depressive symptoms, personality pathology, and (low) life satisfaction and quality are understood, the current research highlights need for additional caution in use of these medications.

4.1. Clinical implications

The current results bolster concerns about routine use of ADM and recommendations to consider other empirically supported treatments, such as cognitive-behavioral therapy, Well-Being Therapy, physical exercise, or even open-label pill placebo [4,61,62]. After potential benefits, harms, and alternatives are discussed openly and fully, if a patient and clinician agree that an ADM trial is in the patient’s best interest, we suggest rigorous, frequent, longitudinal assessment to evaluate the patient’s response and symptom trajectory. Target symptoms might be assessed with the IDAS-II or Patient Health Questionnaire [63], and side effects could be assessed with, for example, the Toronto Side Effect Scale [64] and/or Oxford Questionnaire on the Emotional Side-effects of Antidepressants [65]. Regular assessment both before (e.g., during a run-in period with clinical support and lifestyle changes, patient safety permitting) and after starting ADM would provide information about baseline levels, trend, and variability in symptoms and side effects. For a patient with a symptom decrease that (1) begins following ADM initiation, (2) is meaningfully large (e.g., at least 50% and/or below a clinically relevant threshold [66]), (3) clearly exceeds background variability (e.g., mean symptom decrease is at least 1.5 times the standard deviation of the patient’s series of symptom scores [67], or statistically significant in a formal time-series analysis [68]) and (4) is not overshadowed by troublesome side effects, a harm-benefit analysis may favor short-term ADM use. Even in these cases, however, planning for tapering and stopping ADM within weeks to months (e.g., by switching to cognitive-behavioral therapy or Well-Being Therapy) appears wise to reduce relapse risk, side-effect burden, and discontinuation syndrome [13,69,70].

4.2. Limitations

Several issues in study design and measurement temper our conclusions. First, generalization of the current results to different measures and populations is uncertain (e.g., the current sample included community-dwelling adults at risk for personality pathology). On the other hand, the basic finding of poorer outcomes among adults in the community consuming ADM has now replicated across a range of differently composed and assessed samples [6–11]. Second, observed effect sizes for ADM were small. Without exception, however, all statistically significant relations were in the direction of poorer outcomes among ADM consumers. Thus, the striking finding here is small but potentially harmful effects from a treatment frequently prescribed, consumed, and viewed as helpful. Third, although the current analyses controlled for many relevant covariates, unmeasured confounding variables could account for poorer outcomes among ADM consumers. If so, ADM’s benefits may be reliably observable only under special conditions (e.g., RCTs) that diverge from most community-dwelling adults’ experiences. Finally, our self-report measure did not differentiate ADM type, dose, duration, or adherence to prescription instructions, and receipt of psychotherapy was not assessed. Consequently, our results cannot address which specific patterns of ADM consumption may be more or less helpful or harmful, or how ADM may interact with other interventions.

5. Conclusions

Community-dwelling adults who consumed ADM experienced relative increases in depressive symptomatology and personality pathology, plus decreases in life satisfaction and quality. Thus, consumption of ADM predicted deterioration or worsening on a broad range of psychosocial measures, and in no analysis did ADM consumption predict improvements on measured outcomes. Moderators and/or mechanisms connecting ADM consumption with psychosocial deterioration are unknown, but most associations were independent of demographics, physical health problems, and use of other psychotropic medications. Until the reasons why ADM use increases risk for psychosocial deterioration among community-dwelling adults are better understood, we recommend heightened caution and consideration of empirically supported non-pharmacologic treatments.

Footnotes

Ethics statement

The Institutional Review Board of the University of Notre Dame approved the study procedures (approval number 17-12-4289). All participants provided informed consent. This research was conducted ethically in accordance with the Declaration of Helsinki and the American Psychological Association Ethical Principles of Psychologists and Code of Conduct.

Conflict of interest

Dr. Vittengl is a paid reviewer for UpToDate. Dr. Jarrett is a paid consultant to the National Institutes of Health (NIH), National Institute of Mental Health (NIMH), and UpToDate. Dr. Jarrett has stock equity in Amgen, Johnson and Johnson, and Procter and Gamble. Dr. Ro has no financial interests related to topics covered to declare. Dr. Clark is the author and copyright holder of the Schedule for Nonadaptive and Adaptive Personality-Second Edition. Unfunded, non-commercial research and clinical usage are free of charge, but a license is required. Negotiated fees for commercial-usage licenses and funded, non-commercial usage licenses are used to support students’ research.

Funding

This report was supported by a grant R01-MH083830 to Lee Anna Clark, Ph.D. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIMH or the NIH.

Author contributions

JV conceptualized the research question, analyzed the data, and wrote the paper. RJ conceptualized the research question and wrote the paper. ER collected the data, conceptualized the research question, and wrote the paper. LAC secured the study funding, collected and managed the data, conceptualized the research question, and wrote the paper.

Supplementary materials

The supplementary material is available in the electronic version of this article: .

References

Carl

Witcraft

Kauffman

Psychological and pharmacological treatments for generalized anxiety disorder (GAD): a meta-analysis of randomized controlled trials. Cogn Behav Ther. 2020;49(1):1–21.

Cipriani

Furukawa

Salanti

Comparative efficacy and acceptability of 21 antidepressant drugs for the acute treatment of adults with major depressive disorder: a systematic review and network meta-analysis. Lancet. 2018;391(10128):1357–66.

Jakobsen

Gluud

Kirsch

. Should antidepressants be used for major depressive disorder? BMJ Evid Based Med. 2020;25:130.

Kirsch

. Placebo effect in the treatment of depression and anxiety. Front Psychiatry. 2019;10:407.

Munkholm

Paludan-Müller

Boesen

. Considering the methodological limitations in the evidence base of antidepressants for depression: a reanalysis of a network meta-analysis. BMJ Open. 2019;9(6):e024886.

Fava

. May antidepressant drugs worsen the conditions they are supposed to treat? The clinical foundations of the oppositional model of tolerance. Ther Adv Psychopharmacol. 2020;10:2045125320970325.

Hengartner

Angst

Rössler

. Antidepressant use prospectively relates to a poorer long-rerm outcome of depression: results from a prospective community cohort study over 30 years. Psychother Psychosom. 2018;87(3):181–3.

Hengartner

Passalacqua

Andreae

Antidepressant use during acute inpatient care is associated with an increased risk of psychiatric rehospitalisation over a 12-month follow-up after discharge. Front Psychiatry. 2019;10:79.

Ormel

Hollon

Kessler

Cuijpers

Monroe

. More treatment but no less depression: The treatment-prevalence paradox. Clin Psychol Rev. 2022;91:102111.

10.

Vittengl

. Poorer long-term outcomes among persons with major depressive disorder treated with medication. Psychother Psychosom. 2017;86(5):302–304.

11.

Vittengl

. A mutually-reinforcing system? Longitudinal changes in psychological distress, functional impairment, and prescription medication consumption among adults. Ethical Hum Psychol Psychiatry. 2022;24:112–23.

12.

Mojtabai

Olfson

. National trends in long-term use of antidepressant medications: results from the U.S. National Health and Nutrition Examination Survey. J Clin Psychiatry. 2014;75(2):169–77.

13.

Verhaak

PFM

de Beurs

Spreeuwenberg

. What proportion of initially prescribed antidepressants is still being prescribed chronically after 5 years in general practice? A longitudinal cohort analysis. BMJ Open. 2019;9(2):e024051.

14.

El-Mallakh

Gao

Jeannie Roberts

. Tardive dysphoria: the role of long term antidepressant use in-inducing chronic depression. Med Hypotheses. 2011;76(6):769–73.

15.

Hollon

Andrews

Singla

Maslej

. Mulsant BH. Evolutionary theory and the treatment of depression: It is all about the squids and the sea bass. Behav Res Ther. 2021;143:103849.

16.

Davies

Read

. A systematic review into the incidence, severity and duration of antidepressant withdrawal effects: Are guidelines evidence-based? Addict Behav. 2019;97:111–21.

17.

Horowitz

Framer

Hengartner

Sørensen

Taylor

. Estimating risk of antidepressant withdrawal from a review of published data. CNS Drugs. 2023;37(2):143–57.

18.

Marsden

White

Annand

Medicines associated with dependence or withdrawal: a mixed-methods public health review and national database study in England. Lancet Psychiatry. 2019;6(11):935–50.

19.

Bet

Hugtenburg

Penninx

Hoogendijk

. Side effects of antidepressants during long-term use in a naturalistic setting. Eur Neuropsychopharmacol. 2013;23(11):1443–51.

20.

Carvalho

Sharma

Brunoni

Vieta

Fava

. The safety, tolerability and risks associated with the use of newer generation antidepressant drugs: a critical review of the literature. Psychother Psychosom. 2016;85(5):270–88.

21.

American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders. 5th ed. Washington DC; 2013.

22.

Kotov

Gamez

Schmidt

Watson

. Linking “big” personality traits to anxiety, depressive, and substance use disorders: a meta-analysis. Psychol Bull. 2010;136(5):768–821.

23.

Malouff

Thorsteinsson

Schutte

. The relationship between the five-factor model of personality and symptoms of clinical disorders: a meta-analysis. J Psychopathol Behav Assess. 2005;27(2):101–14.

24.

Sleep

Lynam

Widiger

Crowe

Miller

. An evaluation of DSM-5 Section III personality disorder Criterion A (impairment) in accounting for psychopathology. Psychol Assess. 2019;31(10):1181–91. doi:10.1037/pas0000620.

25.

Vittengl

Jarrett

Clark

. How can the DSM-5 alternative model of personality disorders advance understanding of depression? J Affect Disord. 2023;320:254–62.

26.

Årdal

Lund

Hammar

. Health-related quality of life in recurrent major depressive disorder—a 10-year follow-up study. Nord J Psychiatry. 2013;67(5):339–43.

27.

Gao

Sweet

Calabrese

. Correlation between depression/anxiety symptom severity and quality of life in patients with major depressive disorder or bipolar disorder. J Affect Disord. 2019;244:9–15.

28.

Koivumaa-Honkanen

Kaprio

Honkanen

Viinamäki

Koskenvuo

. Life satisfaction and depression in a 15-year follow-up of healthy adults. Soc Psychiatry Psychiatr Epidemiol. 2004;39(12):994–9.

29.

Clark

Vanderbleek

Shapiro

Nuzum

Allen

Daly

The brave new world of personality disorder-trait specified: Effects of additional definitions on coverage, prevalence, and comorbidity. Psychopathol Rev. 2015;2(1):52–82.

30.

Clark

Nuzum

Shapiro

Personality profiles as potential targets for intervention: Identification and replication. Personal Ment Health. 2020;14(1):142–63.

31.

Langbehn

Pfohl

Reynolds

Clark

Battaglia

Bellodi

The Iowa Personality Disorder Screen: development and preliminary validation of a brief screening interview. J Pers Disord. 1999;13(1):75–89.

32.

Clark

. Three-pronged assessment and diagnosis of personality disorder and its consequences: personality functioning, pathological traits, and psychosocial disability. Personal Disord. 2014;5(1):55–69.

33.

Clark

Watson

. Constructing validity: New developments in creating objective measuring instruments. Psychol Assess. 2019;31(12):1412–27.

34.

Sheehan

Lecrubier

Harnett Sheehan

Janavs

Weiller

Keskiner

The validity of the Mini International Neuropsychiatric Interview (MINI) according to the SCID-P and its reliability. Eur Psychiatry. 1997;12(5):232–41.

35.

Watson

O’Hara

Naragon-Gainey

Koffel

Chmielewski

Kotov

Development and validation of new anxiety and bipolar symptom scales for an expanded version of the IDAS (the IDAS-II). Assessment. 2012;19(4):399–420.

36.

Krueger

Derringer

Markon

Watson

Skodol

. Initial construction of a maladaptive personality trait model and inventory for DSM-5. Psychol Med. 2012;42(9):1879–90.

37.

Clark

Simms

Casillas

. Schedule for Nonadapative and Adaptive Personality—2nd Edition (SNAP-2): Manual for Administration, Scoring, and Interpretation. Notre Dame, Indiana: University of Notre Dame; 2014.

38.

Calabrese

Rudick

Simms

Clark

. Development and validation of Big Four personality scales for the Schedule for Nonadaptive and Adaptive Personality–Second Edition (SNAP-2). Psychol Assess. 2012;24(3):751–63.

39.

Parker

Hadzi-Pavlovic

Both

Kumar

Wilhelm

Olley

. Measuring disordered personality functioning: to love and to work reprised. Acta Psychiatr Scand. 2004;110(3):230–39.

40.

Livesley

. General Assessment of Personality Disorder. Port Huron (MI): Sigma Assessment Systems, Inc; 2010.

41.

Clark

. Interrelations between psychosocial functioning and adaptive- and maladaptive-range personality traits. J Abnorm Psychol. 2013;122(3):822–35.

42.

Diener

Emmons

Larsen

Griffin

. The Satisfaction With Life Scale. J Pers Assess. 1985;49(1):71–5.

43.

WHOQOL Group. Development of the World Health Organization WHOQOL-BREF quality of life assessment. Psychol Med. 1998;28(3):551–8.

44.

Clark

Watson

. Constructing validity: Basic issues in objective scale development. Psychol Assess. 1995;7(3):309–19.

45.

Guadagnoli

Velicer

. Relation of sample size to the stability of component patterns. Psychol Bull. 1988;103(2):265–75.

46.

Cohen

. Statistical Power Analysis for the Behavioral Sciences. 2nd ed. Hillsdale (NJ): Erlbaum; 1988.

47.

Schafer

Graham

. Missing data: our view of the state of the art. Psychol Methods. 2002;7(2):147–77.

48.

Roberts

Walton

Viechtbauer

. Patterns of mean-level change in personality traits across the life course: a meta-analysis of longitudinal studies. Psychol Bull. 2006;132(1):1–25.

49.

Roberts

Luo

Briley

Chow

Hill

. A systematic review of personality trait change through intervention. Psychol Bull. 2017;143(2):117–41.

50.

Soskin

Carl

Alpert

Fava

. Antidepressant effects on emotional temperament: toward a biobehavioral research paradigm for major depressive disorder. CNS Neurosci Ther. 2012;18(6):441–51.

51.

World Health Organization. The WHO Special Initiative for Mental Health (2019–2023): Universal Health Coverage for Mental Health. Geneva, Switzerland: World Health Organization; 2019.

52.

Demyttenaere

Donneau

Albert

Ansseau

Constant

van Heeringen

. What is important in being cured from depression? Discordance between physicians and patients (1). J Affect Disord. 2015;174:390–6.

53.

Fava

Guidi

. The pursuit of euthymia. World Psychiatry. 2020;19(1):40–50.

54.

Hofmann

Curtiss

Carpenter

Kind

. Effect of treatments for depression on quality of life: a meta-analysis. Cogn Behav Ther. 2017;46(4):265–86.

55.

Kamenov

Twomey

Cabello

Prina

Ayuso-Mateos

. The efficacy of psychotherapy, pharmacotherapy and their combination on functioning and quality of life in depression: a meta-analysis. Psychol Med. 2017;47(3):414–25.

56.

Read

Moncrieff

. Depression: why drugs and electricity are not the answer. Psychol Med. 2022;52(8):1401–10.

57.

Andrews

Bharwani

Lee

Fox

Thomson

. Is serotonin an upper or a downer? The evolution of the serotonergic system and its role in depression and the antidepressant response. Neurosci Biobehav Rev. 2015;51:164–88.

58.

Hollon

. Is cognitive therapy enduring or antidepressant medications iatrogenic? Depression as an evolved adaptation. Am Psychol. 2020;75(9):1207–18.

59.

Moncrieff

Cooper

Stockmann

Amendola

Hengartner

Horowitz

. The serotonin theory of depression: a systematic umbrella review of the evidence. Mol Psychiatry. 2022. doi:10.1038/s41380-022-01661-0.

60.

Kirsch

Huedo-Medina

Pigott

Johnson

. Do outcomes of clinical trials resemble those of “real world” patients? A reanalysis of the STARD antidepressant data set. Psychol Conscious. 2018;5(4):339–45.

61.

Guidi

Fava

. The emerging role of euthymia in psychotherapy research and practice. Clin Psychol Rev. 2020;82:101941.

62.

Qaseem

Owens

Etxeandia-Ikobaltzeta

Nonpharmacologic and pharmacologic treatments of adults in the acute phase of major depressive disorder: a living clinical guideline from the American College of Physicians. Ann Intern Med. 2023;176(2):239–52.

63.

Kroenke

Spitzer

Williams

. The PHQ-9: validity of a brief depression severity measure. J Gen Intern Med. 2001;16(9):606–13.

64.

Vanderkooy

Kennedy

Bagby

. Antidepressant side effects in depression patients treated in a naturalistic setting: a study of bupropion, moclobemide, paroxetine, sertraline, and venlafaxine. Can J Psychiatry. 2002;47(2):174–80.

65.

Price

Cole

Doll

Goodwin

. The Oxford Questionnaire on the Emotional Side-effects of Antidepressants (OQuESA): development, validity, reliability and sensitivity to change. J Affect Disord. 2012;140(1):66–74.

66.

Rush

Kraemer

Sackeim

Report by the ACNP Task Force on response and remission in major depressive disorder. Neuropsychopharmacology. 2006;31(9):1841–53.

67.

Kelly

Roberts

Bottonari

. Non-treatment-related sudden gains in depression: the role of self-evaluation. Behav Res Ther. 2007;45(4):737–47.

68.

Barlow

Nock

Hersen

. Single case experimental designs. 3rd ed. New York: Pearson; 2009.

69.

Bockting

ten Doesschate

Spijker

Continuation and maintenance use of antidepressants in recurrent depression. Psychother Psychosom. 2008;77(1):17–26.

70.

Guidi

Fava

. Sequential combination of pharmacotherapy and psychotherapy in major depressive disorder: a systematic review and meta-analysis. JAMA Psychiatry. 2021;78(3):261–9.

	Standardized beta for time 1 ADM:
Time 2 outcome variable	Controlling time 1 scores on the outcome variable	Controlling time 1 scores on the outcome variable plus background covariates^a
Depressive symptom measures
Major depressive disorder symptom count	0.13	0.05
IDAS-II dysphoria	0.12	0.10
IDAS-II well-being	−0.07	−0.07
Personality trait measures
PID-5 negative affect	0.10	0.08
PID-5 detachment	0.07	0.05
PID-5 disinhibition	0.14	0.12
SNAP-2 negative temperament	0.13	0.08
SNAP-2 detachment	0.05	0.05
SNAP-2 conscientiousness	−0.07	−0.10
Personality functioning measures
MDPF non-coping	0.13	0.13
MDPF non-cooperativeness	0.02	0.00
GAPD self-pathology	0.07	0.07
GAPD interpersonal pathology	0.07	0.04
Quality of life measures
Satisfaction with life scale	−0.10	−0.10
WHO quality of life scale	−0.14	−0.11

Associations of antidepressant medication consumption with changes in personality pathology and quality of life among community-dwelling adults

Abstract

BACKGROUND:

OBJECTIVE:

METHOD:

RESULTS:

CONCLUSION:

Keywords

1. Introduction

2. Method

2.1. Participants and recruitment

2.3. Measures

2.3.1. Medication consumption and background variables

2.3.2. Depression

2.3.3. Personality pathology

2.3.4. Quality of life

2.4. Sample size and statistical analyses

4.1. Clinical implications

4.2. Limitations

5. Conclusions

Footnotes

Ethics statement

Conflict of interest

Funding

Author contributions

Supplementary materials

References